Single Cell Transcriptomics of Regulatory T Cells Reveals Trajectories of Tissue Adaptation
Total Page:16
File Type:pdf, Size:1020Kb
bioRxiv preprint doi: https://doi.org/10.1101/217489; this version posted November 22, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Single cell transcriptomics of regulatory T cells reveals trajectories of tissue adaptation Authors: 1,2,a 1,a 3,4 7 Ricardo J Miragaia , Tomás Gomes , Agnieszka Chomka , Laura Jardine , Angela 8 3,4,5 1,9 1 3,4,6 Riedel , Ahmed N. Hegazy , Ida Lindeman , Guy Emerton , Thomas Krausgruber , 8 7 3,4 1,10,11,* Jacqueline Shields , Muzlifah Haniffa , Fiona Powrie , Sarah A. Teichmann Affiliations: 1 Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, UK 2 Centre of Biological Engineering, University of Minho, Braga, Portugal 3 Kennedy Institute of Rheumatology, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK 4 Translational Gastroenterology Unit, Experimental Medicine Division Nuffield Department of Clinical Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK 5 Current affiliation: Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Department of Gastroenterology, Infectious Diseases and Rheumatology 6 Current affiliation: CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna, Austria 7 Institute of Cellular Medicine, Newcastle University, Newcastle-Upon-Tyne, UK 8 MRC Cancer Unit, University of Cambridge, Cambridge, UK 9 Centre for Immune Regulation and Department of Immunology, University of Oslo and Oslo University Hospital, 0372 Oslo, Norway 10 Theory of Condensed Matter, Cavendish Laboratory, Department of Physics, University of Cambridge, Cambridge, UK 11 European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Trust Genome Campus, Hinxton, UK a These authors contributed equally to this work *Lead Contact: [email protected] (S.A.T.) Summary Non-lymphoid tissues (NLTs) harbour a pool of adaptive immune cells, the development and phenotype of which remains largely unexplored. Here, we used single-cell RNA-seq to characterise CD4+ regulatory (Treg) and memory (Tmem) T cells in mouse skin and colon, the respective draining lymph nodes and spleen. From this data, we modelled a continuous lymphoid-to-NLT trajectory for Treg, and reconstructed the mechanisms of cell migration and NLT adaption. This revealed a shared transcriptional programme of NLT priming in both skin and colon-associated lymph nodes, followed by tissue-specific adaptation. Predicted migration kinetics were validated using a melanoma-induction model, emphasizing the relevance of key regulators and receptors, including Batf, Rora, Ccr8, Samsn1. Finally, we profiled human blood and NLT Treg and Tmem cells, identifying cross-mammalian conserved tissue signatures. In summary, we have identified molecular signals mediating 1 bioRxiv preprint doi: https://doi.org/10.1101/217489; this version posted November 22, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. NLT Treg recruitment and tissue adaptation through the combined use of computational prediction and in vivo validation. Introduction CD4+ T cells are major orchestrators of the adaptive immune response controlling host defense towards pathogens as well as maintaining tolerance towards self-antigens. In response to microbial and cytokine cues they differentiate into distinct effector cells tailored to the immune challenge and tissue location. In addition to circulating CD4+ T cells, recent evidence suggests that resident non-lymphoid tissue (NLT) T cell populations play an important role in providing localised immediate immune protection within the NLTs (Fan and Rudensky 2016; Watanabe et al. 2015). Regulatory T cells (Tregs), characterized by Foxp3 expression, are highly specialised cells which control immune responses and play a central role in homeostasis (Sakaguchi 2004; Izcue, Coombes, and Powrie 2009). Multiple studies have described unique tissue-specific adaptations of NLT Tregs distinct from their lymphoid tissue counterparts. This includes acquisition of an effector phenotype with expression of transcripts encoding effector molecules (Ctla4, Gzmb, Klrg1), chemokines and their receptors (e.g. Ccr4), and immunosuppressive cytokines such as Il10 (Panduro, Benoist, and Mathis 2016; Bollrath and Powrie 2013). Homeostatic functions of NLT Tregs include control of insulin resistance in visceral adipose tissue (Feuerer et al. 2009), skeletal muscle repair (Burzyn et al. 2013), and hair follicle stem cell stimulation (Ali et al. 2017). Particular signature genes associated with these and other Treg populations in NLTs have been described (Liston and Gray 2014), but the full extent of their phenotype and heterogeneity within NLTs and how this changes upon challenge has yet to be uncovered. Furthermore, most studies of NLT Tregs have been performed in mice and very little is known about these cells in humans despite the enormous interest in Treg therapy and manipulation in the context of cancer, autoimmune disease, and organ transplant Trafficking of T cells to NLTs occurs in steady-state conditions (Kimpton et al. 1995). For example, Treg were shown to be present in NLTs from a young age (Thome et al. 2015; Cipolletta et al. 2015), showing that the presence of these cells is not dependent on acute NLT immune challenges, and occurs either developmentally or in response to challenges at barrier surfaces from harmless stimuli such as commensal bacteria and dietary antigens (Ivanov et al. 2008). Migration of Tregs to NLTs requires tissue-specific cues involving cell-surface receptors (Chow, Banerjee, and Hickey 2015; Thome et al. 2015), such as chemokine receptors, as well as other G-protein coupled receptors (S. V. Kim et al. 2013) and integrins (Cepek et al. 1994). Despite these insights into trafficking and regulation, characterization of CD4+ T cell 2 bioRxiv preprint doi: https://doi.org/10.1101/217489; this version posted November 22, 2017. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. populations across tissues has historically been held back by phenotypic characterization being restricted to a limited set of markers (e.g. single molecule FISH, FACS) or mixtures of cells (e.g. bulk RNA-sequencing) assayed. To provide further insight into Treg populations in tissues we performed single-cell RNA-seq (scRNA-seq) analyses of Tregs from mouse colon, skin and compared with relevant lymphoid tissues. Our results show NLT Treg adaptations in the steady state with a core shared signature between mouse skin and colon Tregs, indicative of a general NLT residency programme in barrier tissues. Through “pseudospatial” modelling we revealed the transcriptomic adaptations occurring in Tregs during their transition from the lymph node to barrier tissues. These findings were recapitulated during de novo Treg recruitment to melanoma in a murine model system. Lastly, we confirmed the evolutionarily conserved and species-specific expression patterns between mouse and human Tregs. Our expression data for different tissues is also available for user-friendly interactive browsing online at data.teichlab.org. Results + CD4 Treg are transcriptionally distinct across tissues + + + - high We performed scRNA-seq on isolated CD4 F oxp3 (Treg) and CD4 F oxp3 C D44 memory (Tmem) T cells from two barrier NLT site: the colonic lamina propria (hereafter referred to as colon), and the skin, their lymphoid counterparts in the draining mesenteric and brachial lymph nodes (mLN and bLN) and the spleen from a Foxp3-GFP-KI mouse reporter line (Bettelli et al. 2006) (Figure 1A, Supplementary Figure 1). Cells from each NLT (and accompanying lymphoid tissues) were obtained from separate pools of mice, and henceforth we will refer to each as “Colon” or “Skin” datasets. We will refer to Treg and Tmem + populations together as CD4 T cells. A plate-based version of the Smart-seq2 protocol was employed (Picelli et al. 2014), and after mapping and quality control (Supplementary Figure 2, see Methods) we retained 485 cells from the colon dataset, and 796 cells from the skin dataset, expressing on average 3118 genes (TPM>0) (Supplementary Figure 2). We confirmed the expression of